← Back to community benchmarks

qwen3.5-35b-a3b-oQ5e

M5 Max (40c) · 128 GB · 5bit · 2026-04-02
Performance
1k
tokens
1,723
PP tok/s
111.9
TG tok/s
594
TTFT (ms)
23.9
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.0
macOS macOS 26.4
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,723 111.9 23.9 GB current
4k 3,507 107.5 24.8 GB view
8k 3,787 103.1 25.1 GB view
16k 3,630 94.0 25.8 GB view
Batching Results
Batch Size TG tok/s Speedup
111.9 1.00×
182.0 1.63×
264.7 2.37×
337.7 3.02×